EP2558939B1 - Memory controller mapping on-the-fly - Google Patents

Memory controller mapping on-the-fly Download PDF

Info

Publication number
EP2558939B1
EP2558939B1 EP11717074.6A EP11717074A EP2558939B1 EP 2558939 B1 EP2558939 B1 EP 2558939B1 EP 11717074 A EP11717074 A EP 11717074A EP 2558939 B1 EP2558939 B1 EP 2558939B1
Authority
EP
European Patent Office
Prior art keywords
memory
banks
bank
memory banks
addresses
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
EP11717074.6A
Other languages
German (de)
English (en)
French (fr)
Other versions
EP2558939A1 (en
Inventor
Ian Hendry
Rajabali Koduri
Jeffry Gonion
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Apple Inc
Original Assignee
Apple Inc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Apple Inc filed Critical Apple Inc
Publication of EP2558939A1 publication Critical patent/EP2558939A1/en
Application granted granted Critical
Publication of EP2558939B1 publication Critical patent/EP2558939B1/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0646—Horizontal data movement in storage systems, i.e. moving data in between storage devices or systems
    • G06F3/0647—Migration mechanisms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3206—Monitoring of events, devices or parameters that trigger a change in power modality
    • G06F1/3215—Monitoring of peripheral devices
    • G06F1/3225—Monitoring of peripheral devices of memory devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F1/00—Details not covered by groups G06F3/00 - G06F13/00 and G06F21/00
    • G06F1/26—Power supply means, e.g. regulation thereof
    • G06F1/32—Means for saving power
    • G06F1/3203—Power management, i.e. event-based initiation of a power-saving mode
    • G06F1/3234—Power saving characterised by the action undertaken
    • G06F1/325—Power saving in peripheral device
    • G06F1/3275—Power saving in memory, e.g. RAM, cache
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/0223—User address space allocation, e.g. contiguous or non contiguous base addressing
    • G06F12/0292—User address space allocation, e.g. contiguous or non contiguous base addressing using tables or multilevel address translation means
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/06—Addressing a physical block of locations, e.g. base addressing, module addressing, memory dedication
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/06—Addressing a physical block of locations, e.g. base addressing, module addressing, memory dedication
    • G06F12/0646—Configuration or reconfiguration
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F12/00—Accessing, addressing or allocating within memory systems or architectures
    • G06F12/02—Addressing or allocation; Relocation
    • G06F12/08—Addressing or allocation; Relocation in hierarchically structured memory systems, e.g. virtual memory systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/10—File systems; File servers
    • G06F16/17—Details of further file system functions
    • G06F16/1737—Details of further file system functions for reducing power consumption or coping with limited storage space, e.g. in mobile devices
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/0604—Improving or facilitating administration, e.g. storage management
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
    • G06F3/0625—Power saving in storage systems
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
    • G06F3/0646—Horizontal data movement in storage systems, i.e. moving data in between storage devices or systems
    • G06F3/065—Replication mechanisms
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F3/0601—Interfaces specially adapted for storage systems
    • G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
    • G06F3/0671—In-line storage system
    • G06F3/0673—Single storage device
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
    • G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
    • G06F2003/0697—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers device management, e.g. handlers, drivers, I/O schedulers
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/50—Reducing energy consumption in communication networks in wire-line communication networks, e.g. low power modes or reduced link rate

Definitions

  • the present disclosure relates generally to memory management techniques and, more particularly, to techniques for memory mapping on-the-fly.
  • US7100013 relates to a method and apparatus to reduce the power consumption of portable devices to increase battery life by turning off power to at least a subset of memory associated with the portable devices.
  • the power consumption is reduced by minimizing the number of usable memory locations.
  • an electronic device may include several memory banks, one or more processors, and a memory controller.
  • the memory banks may store data in hardware memory locations and may be independently deactivated.
  • the processors may request the data using physical memory addresses, and the memory controller may translate the physical addresses to the hardware memory locations.
  • the memory controller may use a first memory mapping function when a first number of memory banks is active and a second memory mapping function when a second number is active.
  • the memory controller may copy data from only the memory bank that is to be deactivated to the active remainder of memory banks.
  • Present embodiments relate to power and memory management for an electronic device.
  • the present disclosure describes techniques for memory mapping and remapping on-the-fly and for powering up or down one or more portions of total memory, which may be, in certain embodiments, one or more memory banks.
  • the terms “power down”, “shut down”, and “deactivate” refer to placing memory into any low power condition, such as turning off the memory, placing the memory in a self-refresh mode, or setting the memory to any other lower-power-consumption mode.
  • the presently-disclosed techniques describe a manner of on-the-fly memory remapping when a portion of the available memory is powered down using certain disclosed mapping functions.
  • an electronic device may have three banks of memory, each of which may have a particular memory bus.
  • a memory controller may map certain physical addresses to hardware memory locations (also referred to herein as “storage cells” or “dual in-line memory module (DIMM) addresses”) distributed approximately evenly across all three memory banks.
  • DIMM dual in-line memory module
  • physical address refers to a memory chunk that is manipulable by the memory controller and may be any suitable size.
  • the term "physical address” is used in reference to mapping or remapping memory, the term may refer to cache lines or pages of memory that are mapped or remapped, even if the memory controller is capable of manipulating smaller chunks.
  • the term "physical address,” when used in reference to the disclosure herein, may refer to the remapping of pages of memory, even though cache lines of the pages may be individually accessible through the memory controller.
  • the data When data associated with the physical addresses is accessed, the data may be transmitted with a maximized bandwidth over all three memory buses at once.
  • one or more of the memory banks and/or buses may be powered down (e.g., turned off, placed in a self-refresh mode, placed in a lower power consumption mode, etc.).
  • the memory controller may remap, on-the-fly, the remaining physical addresses to DIMM addresses on the first two memory banks via a mathematical memory mapping function.
  • the memory controller may copy data associated with the remapped physical addresses directly from the third memory bank to the first two memory banks.
  • no data may be copied from the first memory bank to the second memory bank, or from the second memory bank to the first memory bank, during the remapping process.
  • the memory controller may cause the third memory bank and/or memory bus to be shut down. It should be understood that the present techniques may also be applied to any number of memory banks and/or portions of the total memory.
  • FIG. 1 a block diagram depicting various components that may be present in electronic devices suitable for use with the present techniques is provided.
  • FIG. 2 one example of a suitable electronic device, here provided as a notebook computer system, is depicted. These types of electronic devices, and other electronic devices having comparable memory management capabilities, may be used in conjunction with the present techniques.
  • FIG. 1 is a block diagram illustrating various components and features of an electronic device 10 capable of performing the techniques disclosed herein.
  • such components may include one or more processor(s) 12, memory 14, nonvolatile storage 16, a display 18, input structures 20, input/output (I/O) ports 22, a networking device 24, and a power source 26.
  • the various functional blocks shown in FIG. 1 may include hardware elements (including circuitry), software elements (including computer code stored on a non-transitory computer-readable medium) or a combination of both hardware and software elements.
  • FIG. 1 is merely one example of a particular implementation and is intended to illustrate the types of components that may be present in electronic device 10.
  • the processor(s) 12 may enable the operation of an operating system (OS), which itself may enable the use of various software to temporarily store information in the volatile memory 14.
  • OS operating system
  • the OS running on the processor(s) 12 may operate using logical memory addresses that may be translated by the processor(s) 12 into physical addresses, and the processor(s) 12 may request certain memory operations by the memory 14 based on these physical addresses.
  • a memory controller of the memory 14 may translate the physical addresses to hardware memory locations of any suitable size (e.g., lines or pages), such as dual inline memory module (DIMM) addresses, for example.
  • DIMM dual inline memory module
  • the DIMM addresses may represent the actual, physical location of the stored data to which the physical addresses, and thus also the logical addresses, correspond. Also as discussed in greater detail below, the memory 14 may be capable of power conservation by reducing the number of operational DIMMs by approximately one-third or more and remapping, on-the-fly, the correspondence of the remaining physical addresses to the remaining DIMM addresses. The number of operational DIMMs may be reduced further, in which case the remaining physical addresses also may be remapped to the remaining DIMM addresses.
  • the memory 14 may store instructions for carrying out certain aspects of the present techniques described herein.
  • the OS running on the processors 12 and/or software running on the OS may perform certain algorithms relating to the present techniques (e.g., when to enter a low-power mode, etc.).
  • the instructions for carrying out these instructions may be stored, at least temporarily, in the memory 14.
  • Such instructions may also be stored in the nonvolatile storage 16, which may include, for example, a hard disk drive or Flash memory.
  • the display 18 may display elements of the OS or software, such as the user interface (UI) of the electronic device 10.
  • UI user interface
  • a user may interact with the electronic device 10 via the input structures 20, which may include a keyboard and/or a mouse or touchpad.
  • the display 18 may be a touchscreen display that serves as one of the input structures 20.
  • the I/O ports 22 of the electronic device 10 may enable the electronic device 10 to transmit data to and receive data from other electronic devices 10 and/or various peripheral devices, such as external keyboards or mice.
  • the networking device 24 may enable personal area network (PAN) integration (e.g., Bluetooth), local area network (LAN) integration (e.g., Wi-Fi), and/or wide area network (WAN) integration (e.g., cellular 3G or 4G).
  • PAN personal area network
  • LAN local area network
  • WAN wide area network
  • the power source 26 of the electronic device 10 may be any suitable source of power, such as a rechargeable lithium polymer (Li-poly) battery and/or standard alternating current (AC) power converter.
  • Li-poly rechargeable lithium polymer
  • AC alternating current
  • the electronic device 10 may take the form of a computer or other type of electronic device. Such computers may include computers that are generally portable (such as laptop, notebook, and tablet computers) as well as computers that are generally used in one place (such as conventional desktop computers, workstations and/or servers).
  • the electronic device 10 in the form of a computer may be a model of a MacBook®, MacBook® Pro, MacBook Air®, iMac®, Mac® mini, or Mac Pro® available from Apple Inc.
  • the electronic device 10 may take the form of a handheld or tablet device, such as a model of an iPod®, iPhone®, or iPad® available from Apple Inc.
  • the electronic device 10, taking the form of a notebook computer 28, is illustrated in FIG. 2 in accordance with one embodiment of the present disclosure.
  • the depicted computer 28 may include a housing 30, a display 18, input structures 20, and I/O ports 22.
  • the input structures 20 (such as a keyboard and/or touchpad) may be used to interact with the computer 28, such as to start, control, or operate a GUI or applications running on computer 28.
  • a keyboard and/or touchpad may allow a user to navigate a user interface or application interface displayed on the display 18.
  • the electronic device 10 may be capable of displaying high-performance graphics on the display 18, which may involve accessing large quantities of the memory 14 at high bandwidths at certain times. However, at other times, the electronic device 10 may not significantly benefit from such large amounts of the memory 14 or such high bandwidths.
  • the electronic device 10 may include a memory management system 34, illustrated in FIG. 3 , which may enable certain parts of the memory 14 to be shut down (e.g., turned off, placed in a self-refresh mode, placed in a lower power consumption mode, etc.) to conserve power.
  • the processor(s) 12 may communicate with the memory 14 through a memory controller 36 and a bus 38.
  • an operating system (OS) running on the processor(s) 12 may allocate and deallocate memory at logical addresses.
  • a memory management unit (MMU) 39 of the processor(s) 12 may translate the logical addresses employed by the OS into physical addresses that generally correspond to a memory addresses understood by the memory controller 36.
  • the MMU 39 may perform this translation using one or more translation look aside buffers (TLB) or any other suitable hardware.
  • the OS of the processor(s) 12 may translate the logical addresses to the physical addresses.
  • the memory controller 36 may receive instructions to read data from or write data to one of the physical address from the processor(s) 12.
  • Contiguous physical addresses requested by the processor(s) 12 may not directly correspond to contiguous hardware memory locations located on memory banks 40A, 40B, and 40C, shown in FIG. 3 as bank 0, bank 1, and bank 2. Rather, the memory controller 36 may map the physical addresses requested by the processor(s) 12 to certain hardware memory locations, also referred to herein as “storage cells” or “dual in-line memory module (DIMM) addresses,” that may be distributed approximately evenly across all of the memory banks 40A, 40B, and 40C.
  • storage cells also referred to herein as “storage cells” or “dual in-line memory module (DIMM) addresses
  • the memory banks 40A, 40B, and 40C may include any suitable memory devices, such as double data rate three synchronous dynamic random access memory (DDR3 SDRAM), double data rate three synchronous dynamic random access memory (DDR4 SDRAM), and/or graphics double data rate five synchronous dynamic random access memory (GDDR5 SDRAM).
  • the memory controller 36 may communicate with the three memory banks 40A, 40B, and 40C via memory buses 42, 44, and 46, which may respectively interconnect the memory controller 36 with banks 0, 1, and 2.
  • the memory management system 34 may include more or fewer memory buses and memory banks 40.
  • the physical addresses may not be contiguously mapped across the DIMM addresses memory banks 40A, 40B, and 40C, but rather may be mapped such that no two "related" or “contiguous” physical addresses are mapped to the same memory bank 40.
  • the terms "related" or “contiguous” physical addresses refer to physical addresses that are likely to be accessed in succession (e.g., physical addresses mapped to contiguous software logical memory addresses by the OS), even if the physical addresses are not strictly numerically contiguous.
  • a first physical address may be mapped to a DIMM address on bank 0
  • a second physical address may be mapped to a DIMM address on bank 1, and so forth.
  • This scheme may be altered when the memory banks 40A, 40B, and 40C have different bandwidth characteristics, such that a higher bandwidth memory bank may have mapped to it more than one related physical address.
  • some embodiments may involve mapping the first and second physical addresses to DIMM addresses on bank 0 and bank 1, respectively, and mapping third and fourth physical addresses to bank 2.
  • the physical addresses may be distributed to DIMM addresses of the memory banks 40A, 40B, and/or 40C to reduce latency "low points.”
  • the memory controller 36 may, in certain embodiments, gain access to three physical addresses at a time via the three memory buses 42, 44, and 46, which may effectively maximize the data transfer bandwidth.
  • the memory controller 36 generally may map the physical addresses requested by the processor(s) 12 to the DIMM addresses of the memory banks 40A, 40B, and 40C such that each memory bank 40 includes as many related physical addresses as possible without incurring undue latency. Therefore, if each memory bank 40 is capable of providing access only to one physical address at a time without incurring additional latency, the physical addresses may be distributed such that no two related physical addresses map to two DIMM addresses on the same memory bank 40, as discussed above. If each memory bank 40 is capable of providing access to no more than two physical addresses at a time without incurring additional latency, the physical addresses may be distributed such that no three related physical addresses map to three DIMM addresses on the same memory bank 40.
  • the memory controller 36 may not map the physical addresses requested by the processor(s) 12 to the DIMM addresses of the memory banks 40A, 40B, and 40C such that each memory bank 40 includes as many related physical addresses as possible without incurring undue latency, as discussed briefly above. Instead, the memory controller 36 may map the physical addresses to the DIMM addresses such that each memory bank 40 includes fewer related physical addresses than the maximum possible without latency. Thereafter, when the memory controller 36 causes a memory bank 40 and/or memory bus 42, 44, and/or 46 to be shut down to conserve power, remapping the remaining physical addresses to the DIMM addresses of the remaining memory banks 40, additional latency may not be incurred even if two related physical addresses are mapped to the same memory bank 40.
  • each memory bank 40 is capable of providing access to no more than two physical addresses at a time without incurring additional latency
  • the physical addresses may be distributed such that no two related physical addresses map to three DIMM addresses on the same memory bank 40.
  • the memory controller 36 reduces the number of active memory banks 40A, 40B, and 40C and two related physical addresses then may be mapped to the same memory bank 40, the memory bank 40 may not incur additional latency.
  • not all memory buses 42, 44, and 46 may have the same bandwidth capabilities, and not all memory banks 40A, 40B, and 40C may operate with the same latency. Indeed, the memory banks 40A, 40B, and 40C may have different signaling characteristics or different sizes.
  • at least one memory bank 40 may be DDR memory, while another memory bank 40 may be Flash memory.
  • the memory controller 36 may account for such asymmetries by mapping the physical addresses to the DIMM addresses of the memory banks 40A, 40B, and 40C such that each includes as many related physical addresses as possible (or fewer, as in the alternative embodiments mentioned above) without incurring undue latency, though the mapped distribution of physical addresses to DIMM addresses may not evenly be distributed.
  • the memory controller 36 may control when one or more of the memory banks 40A, 40B, and 40C and/or memory buses 42, 44, and 46 are shut down (e.g., by turning off the memory, placing the memory in a self-refresh mode, or setting the memory to any other lower-power-consumption mode) to conserve power.
  • the memory controller 36 undertakes steps to shut down a memory bank 40A, 40B, or 40C and/or one of the memory buses 42, 44, and/or 46, the memory controller 36 may dynamically remap a reduced set of the physical addresses to the DIMM addresses of the remaining memory banks 40 on the fly.
  • FIG. 4 illustrates one embodiment of such a dynamic mapping scheme.
  • a memory mapping diagram 48 illustrates a manner in which related physical addresses may be distributed approximately evenly across the memory banks 40A, 40B, and 40C, according to a dynamic mapping function, or equations, employed by the memory controller 36.
  • the memory controller 36 may employ any mapping function that dynamically maps physical addresses to DIMM addresses across the three memory banks 40A, 40B, and 40C in one of the manners described above.
  • mapping physical addresses to the certain hardware memory locations, it should be understood that other memory mapping configurations may involve distributing pages or any other suitable granularities of memory across multiple memory banks 40.
  • a leftmost column labeled "physical address” indicates physical addresses that may be requested of the memory controller 36 by the processor(s) 12.
  • a column labeled “bank” lists each memory bank 40A, 40B, and 40C (bank 0, bank 1, and bank 2). For explanatory purposes only, each memory bank 40 holds eight memory addresses. In practice, it should be understood that the size of the memory banks 40 and their memory addresses may be much larger, and may represent any size chunk of memory that may be manipulated by the memory controller 36 (e.g., a cache line or page).
  • a column labeled "DIMM address” illustrates the mapping of the physical addresses to certain hardware memory locations, or DIMM addresses, on the three memory banks 40A, 40B, and 40C.
  • the DIMM address 0 may hold data associated with the physical address
  • the DIMM address 1 may hold data associated with the physical address 3, and so forth.
  • the physical addresses may not be directly, contiguously mapped to the DIMM addresses.
  • the physical addresses may be understood to correspond sequentially to a quantity of memory addresses that each memory bank 40 could hold. That is, the first eight physical addresses are noted via a first hatching, the second eight physical addresses are noted via a second hatching, and the third eight physical addresses are noted via a third hatching. These hatchings are intended to more clearly illustrate how the memory controller 36 may map and remap these physical addresses to DIMM addresses across the memory banks 40A, 40B, and 40C.
  • the memory controller 36 may dynamically map the physical address to its corresponding DIMM address in the memory banks 40A, 40B, and 40C according to a memory mapping function.
  • the variable PAddr represents the physical address listed in the leftmost column of the memory mapping diagram 48 and the variable DIMMAddr represents the DIMM address to which the physical address is mapped.
  • the variable Banklndex represents an offset value for varying to which of the DIMM addresses of a memory bank 40 a physical address is mapped.
  • the variable BankSize represents the quantity of physical addresses per memory bank 40.
  • the variable TotalBanks represents the total number of memory banks 40A, 40B, and 40C and the variable TotalMemory represents the total number of DIMM addresses. This may be referred to as strip memory, which wraps around, but offsets, physical addresses by their nominal original bank to avoid sequentially mapping to the same DIMM address.
  • the memory controller 36 may map the physical address "4" to the DIMM address "9" as follows: MOD PAddr ⁇ BankSize , TotalMemory + BankIndex MOD 4 ⁇ 8,24 + Floor 4 / 3 MOD 32,24 + 1 8 + 1 9
  • the memory controller 36 may map the physical address "2" to the DIMM address "16" as follows: MOD PAddr ⁇ BankSize , TotalMemory + BankIndex MOD 2 ⁇ 8,24 + Floor 2 / 3 MOD 16,24 + 0 16 + 0 16
  • the memory controller 36 may dynamically map the physical addresses to DIMM addresses in accordance with the mapping function described by Equations (1) and (2).
  • the memory mapping diagram 48 based on the memory mapping function of Equations (1) and (2), physical addresses may be mapped to DIMM addresses in a manner that prevents sequential, related physical addresses from being mapped to the same memory bank 40. This configuration may generally maximize the bandwidth available when the processor(s) 12 request a series of sequential physical addresses when the memory banks are capable of accessing one DIMM addresses at a time without latency.
  • the mapping function may be adapted to distribute the physical addresses such that two or more sequential physical addresses map to DIMM addresses on the same memory bank 40.
  • the memory controller 36 may still employ the mapping function described above. In so doing, when the memory controller 36 causes one of the memory banks 40A, 40B, and 40C and/or one of the memory buses 42, 44, and 46 to be shut down to conserve power, and remaps the remaining physical addresses to the DIMM addresses of the remaining memory banks 40A, 40B, and 40C, latency may not significantly increase, even when two related physical addresses are subsequently mapped to the same memory bank 40.
  • the electronic device 10 may use various quantities of the memory 14 at different points in time.
  • memory associated with graphics may be very volatile, used in highly variable quantities at different times.
  • FIGS. 5 and 6 represent schematic diagrams illustrating such variations in memory usage.
  • a memory usage diagram 50 represents the total amount of memory available across the three memory banks 40A, 40B, and 40C.
  • In-use memory 52 may store data associated with, among other things, a screen on the display 18, a cursor, windows that may have animation, assembly buffers, and so forth.
  • the in-use memory 52 may not occupy all of the total available memory, however, as a certain amount of unused memory 54 may remain in the memory banks 40A, 40B, and 40C.
  • the total amount of memory currently in active or frequent use by the electronic device 10 may span more than the total quantity of memory available to two of the memory banks 40A, 40B, and 40C. As such, all three memory banks 40A, 40B, and 40C may be employed.
  • the memory mapping function employed by the memory controller 36 may map physical addresses to DIMM addresses located across the three memory banks 40A, 40B, and 40C.
  • a memory usage diagram 56 may include a smaller amount of in-use memory 52 and a larger amount of unused memory 54.
  • the amount of in-use memory 52 may span more than one memory bank 40 but less than all three memory banks 40A, 40B, and 40C.
  • a memory management factor diagram 58 of FIG. 7 illustrates several such device memory operation criteria 60, which may include, among other things, indications by performance counters 62, an operational state 64 of the electronic device 10, a memory usage history 66 of the electronic device 10, an expected memory usage 68 of the electronic device 10, thermal limitations 70 of the electronic device 10, GPU memory stalls 72, paging activities 74, and/or user preferences 76.
  • the performance counters 62 may represent a continuous monitoring by the operating system (OS) of the system performance of the electronic device 10. For example, the performance counters 62 may indicate how much processing is taking place at a given time and/or may provide a recent history of how system resources have been consumed. A higher consumption of processing resources, as indicated by the performance counters 62, may signal that the available memory resources should be increased, if possible. Similarly, a reduced consumption of processing resources may signal that the available memory resources should be reduced, if possible.
  • OS operating system
  • the operational state 64 may be determined based on the performance counters 62, and may represent, for example, whether the electronic device 10 is operating in an active, high performance, or idle state. Additionally or alternatively, the operational state 64 may indicate whether the electronic device 10 is operating in a reduced power state, for example, because the power source 26 has changed from an external AC source to a battery source or because battery power is low. Certain operational states 64 (e.g., an active state) may signal that the available memory resources should be increased, if possible. Other operational states 64 (e.g., an idle state) may signal that the available memory resources should be reduced, if possible.
  • the memory usage history 66 and expected memory usage 68 may represent criteria for determining whether more or less memory should be made available to the electronic device 10.
  • the memory usage history 66 may represent recent historical memory usage (e.g., memory usage of the prior 1 minute, 2 minutes, 5 minutes, 10 minutes, 20 minutes, 1 hour, etc.) and/or may represent memory usage patterns for extended periods of time (e.g., memory usage over prior days, weeks, months, and/or the life of the electronic device 10).
  • Expected memory usage 68 may derive from the memory usage history 66 or may be based on typical patterns of memory use for the electronic device 10.
  • the memory usage history 66 and/or the expected memory usage 68 may signal whether the available memory resources should be increased or reduced, depending on whether more or less memory is in use or expected to be in use.
  • the memory usage history 66 and/or the expected memory usage 68 may be illustrated by the following examples.
  • the operating system (OS) of the electronic device 10 may recognize when certain application programs with known memory consumption histories are launched. When an application that has historically consumed large quantities of memory is launched, the OS or the application may provide an indication that a relatively large quantity of memory may be consumed in the near future.
  • the electronic device 10 may predict when large quantities of memory are typically needed for use by the user based on the user's historical usage patterns. For example, the electronic device 10 may recognize that on Tuesday morning the electronic device 10 typically performs a significant quantity of image processing, or that on Thursday evening, the electronic device 10 is used to play memory-intense 3-D games.
  • the memory usage history 66 and the expected memory usage 68 may weigh significantly in favor of increasing available memory resources.
  • an application currently running on the electronic device 10 may provide a cue to indicate that a relatively large amount of memory is about to be needed in the near future or that a relatively low amount of memory is about to be needed in the near future. Based on such cues, the expected memory usage 68 (e.g., as determined by the OS of the electronic device 10) may weigh in favor of increasing or decreasing the currently available memory as needed.
  • the OS may follow the allocation patterns of memory to determine when large patterns of memory allocation have occurred in the past and may be likely to occur in the future.
  • thermal limitations 70 that are placed on the electronic device 10. Due to external heat from the environment and internal heat generated by components of the electronic device 10, particularly after extended periods of high performance operation, the electronic device 10 may approach the thermal limitations 70. If such thermal limitations 70 are approached, the electronic device 10 may be more likely to determine that the available memory resources should be reduced. With reduced memory resources generating heat, the amount of internally-generated heat may accordingly be reduced and the electronic device 10 may retreat from the thermal limitations 70.
  • device memory stalls such as CPU or GPU memory stalls 72, or memory stalls by any other data processing circuitry of the electronic device 10
  • memory paging activities 74 may indicate whether more or less memory should be made available to the electronic device 10.
  • the presence of GPU memory stalls 72 may indicate that more memory should be made available, while the lack of such GPU memory stalls 72 may indicate that the electronic device 10 currently has an excess supply of memory.
  • memory paging activities 74 may indicate a degree to which the available memory is or is not in active or frequent use.
  • a further criterion that may indicate whether to increase or decrease the available memory resources of the electronic device 10 may be user preferences 76.
  • the user of the electronic device 10 may indicate a preference for increased battery life over high performance, or vice versa.
  • the user may elect to cause the electronic device 10 to enter a lower or higher memory consumption mode.
  • the electronic device 10 may be capable of use by multiple users, and different users may operate the electronic device 10 in different ways. That is, some users may generally employ memory-intense applications, while others may generally employ applications that consume relatively little memory.
  • the electronic device 10 may determine each user's preferences 76. Thereafter, when the user is using the electronic device 10, the user preferences 76 may weigh in favor of increasing or decreasing the available memory.
  • the electronic device 10 may consider the memory operation criteria 60 in determining how to manage the memory 14 of the electronic device 10.
  • One embodiment of a method for managing the memory 14 based on such criteria 60 appears in a flowchart 80 of FIG. 8 .
  • the electronic device 10 e.g., the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS
  • the criteria 60 may be weighed against one another to ascertain whether, on balance, electronic device 10 memory usage is excessive or insufficient, or is expected to become excessive or insufficient in the future.
  • certain criteria 60 may be given different weights and/or certain criteria 60 may override all others (e.g., thermal limitations 70 may receive priority when heat becomes excessive).
  • the electronic device 10 may consider whether the excessiveness of the memory usage has crossed a threshold justifying the deactivation of one or more portions of the memory 14 in decision block 86.
  • a threshold may include a time component, which may represent a length of time that the memory usage has been excessive, and/or may include a quantitative component, which may represent the extent to which the memory usage is deemed in excess.
  • the threshold may represent an amount of total excess memory usage and/or an amount of time that the amount of in-use memory has remained below a certain number of portions of the memory 14 (e.g., usage of less than 2 memory banks 40, as shown in FIG. 6 ). If the threshold has not been reached in decision block 86, the electronic device 10 may reduce some power consumption by reducing the clock speed and/or operating voltage of the memory banks 40A, 40B, and/or 40C and/or their associated memory buses 42, 44, and/or 46 in block 88 to place them into a reduced power consumption mode.
  • the electronic device 10 may reduce additional power consumption by deactivating one of the memory banks 40A, 40B, and/or 40C and/or their associated memory buses 42, 44, and/or 46 in block 90. A more detailed description of how such deactivation may take place is described below.
  • decision blocks 84 and 86 to block 90 when one or more memory banks 40 are powered down, may be carried out in a variety of other manners.
  • logic associated with the memory controller 36 may initiate the shut-down process when the memory usage of the electronic device 10 has been reduced to a total number of physical addresses that may be mapped to DIMM addresses in a fewer number of memory banks 40A, 40B, and 40C than currently in use.
  • the operating system (OS) running on the processor(s) 12, or software running on the OS may follow decision blocks 84 and 86 above to determine when to power down a memory bank 40.
  • the OS may then instruct the processor(s) 12 to send a control signal to the memory controller 36, causing the shut-down process to begin in block 90.
  • the processor(s) 12 may automatically detect when the memory usage has fallen to encompass fewer memory banks 40A, 40B, and 40C than currently in use. For example, the MMU 39 may periodically or continuously provide to the processor(s) 12 an indication of the total number of logical addresses that are currently in use and/or mapped to physical addresses. Thereafter, the processor(s) 12 may send a control signal to the memory controller 36 to initiate the shut-down process.
  • logic associated with the memory controller 36 may automatically detect when fewer memory banks 40A, 40B, and 40C should be used by monitoring which physical addresses are being requested by the processor(s) 12. For example, if no physical addresses extending beyond bank 1 are requested by the processor(s) 12 for a threshold period of time (e.g., 10 seconds, 20 seconds, 1 minute, 2 minutes, 5 minutes, 10 minutes, 20 minutes, etc.), the memory controller 36 may understand that these physical addresses are not needed, and may thus initiate the shut-down process automatically.
  • a threshold period of time e.g. 10 seconds, 20 seconds, 1 minute, 2 minutes, 5 minutes, 10 minutes, 20 minutes, etc.
  • the electronic device 10 may determine, in decision block 92, whether the total memory currently available is insufficient in light of the memory operation criteria 60.
  • the electronic device 10 may consider whether the insufficiency of the memory usage has crossed a threshold justifying the reactivation of one or more portions of the memory 14 that have previously been shut down.
  • a threshold may include a time component, which may represent a length of time that the memory usage has been insufficient, and/or may include a quantitative component, which may represent the extent to which the memory usage is deemed insufficient.
  • the threshold may represent an amount of total memory usage and/or an amount of time that the amount of in-use memory has remained above a limit approaching a certain number of portions of the memory 14 (e.g., usage of nearly all of 2 memory banks 40, when one memory bank 40 has been shut down).
  • the performance of the electronic device 10 may be increased, without significantly increasing power consumption, by increasing the clock speed and/or operating voltage of the memory banks 40A, 40B, and/or 40C and/or their associated memory buses 42, 44, and/or 46 in block 96. If the threshold has been reached in decision block 94, the electronic device 10 may reactivate one of the memory banks 40A, 40B, and/or 40C and/or their associated memory buses 42, 44, and/or 46, which previously may have been shut down, in block 98. If the criteria 60 indicate neither that the memory usage is excessive in decision block 84, nor that the memory usage is insufficient in block 92, the electronic device 10 may not substantially change the configuration of the memory 14, as indicated by block 100.
  • FIG. 9 illustrates a manner by which the memory management system 34 may conserve power by causing the memory bank 40C (bank 2) and/or the associated memory bus 46 to be shut down, after remapping and copying certain physical addresses from the memory bank 40C to the remaining memory banks 40A and 40B.
  • the memory controller 36 is illustrated as having caused memory bank 40C (bank 2) and the associated memory bus 46 to shut down.
  • the memory controller 36 may alternatively cause any other of the memory banks 40A or 40B and/or memory buses 42 or 44 to be shut down, depending on design considerations and/or an operational status of the electronic device 10. Moreover, while one entire memory bank 40C is illustrated as having been shut down, in some embodiments, the memory controller 36 may cause only a portion of the memory bank 40C to be shut down.
  • the memory controller 36 may cause the selected memory bank 40C and/or memory bus 46 to shut down in any suitable manner.
  • the memory bank 40C may be capable of shutting down upon the receipt of a specific control signal or instruction from the memory controller 36.
  • the memory controller 36 may cause a power supply to the selected memory bank 40 to be cut off by, for example, directing a switch (not shown) to cut off power.
  • the memory controller 36 may begin the process of powering down one of the memory banks 40A, 40B, or 40C and/or a memory bus 42, 44, or 46.
  • the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS may determine whether to shut down one of the memory banks 40A, 40B, or 40C based on one or more of the memory operation criteria 60.
  • block 112 may occur after the OS determines that its working set of memory is or could be made small enough not to use memory from one of the memory banks 40A, 40B, or 40C.
  • the OS then may stop using an amount of memory equal to that of one of the memory banks 40A, 40B, or 40C. For example, if each of the memory banks 40A, 40B, and 40C have the same capacity, the OS may stop using one-third of the memory.
  • the memory controller 36 may select which of the memory banks 40A, 40B, or 40C to shut down in block 114.
  • the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS may select for shutdown one of the memory banks 40A, 40B, and 40C in a certain order. For example, if all memory banks 40A, 40B, and 40C are currently active, bank 2 may be selected for shut-down; if only two of the memory banks 40A, 40B, and 40C are currently active, bank 1 may be selected for shut-down.
  • the OS may call the memory controller 36 to cause the memory bank 40A, 40B, or 40C to be shut down (e.g., turned off or placed into a self-refresh mode).
  • the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS may determine which memory bank 40 and/or memory bus 42, 44, or 46 to shut down based on power consumption factors balanced with bandwidth factors. For example, certain of the memory banks 40A, 40B, and 40C and/or memory buses 42, 44, and 46 may have higher bandwidth capabilities, but may consume more power. When bandwidth concerns outweigh power consumption concerns, the memory controller 36 may select a lower-bandwidth memory bank 40 and/or memory bus 42, 44, or 46 to shut down, conserving some power but preserving a greater amount of memory bandwidth. When power consumption concerns outweigh bandwidth concerns, the memory controller 36 may instead select a higher-bandwidth memory bank 40 and/or memory bus 42, 44, or 46 to shut down.
  • the memory controller 36 may remap the remaining, or active, physical addresses to DIMM addresses and copy the data from the memory bank 40 selected to be shut down prior to shutting down the selected memory bank 40.
  • the memory controller 36 may cause the selected memory bank 40 and/or memory bus 42, 44, or 46 to shut down in any suitable manner.
  • the memory bank 40 may capable of shutting down upon the receipt of a specific control signal or instruction from the memory controller 36.
  • the memory controller 36 may cause a power to the selected memory bank 40 to be cut off by, for example, directing a switch (not shown) to cut off power.
  • a memory mapping diagram 120 illustrates a manner of dynamically remapping from the three-bank mapping scheme to a two-bank mapping scheme on the fly, as employed during block 116 of FIG. 10 .
  • a column labeled "Physical Address” lists physical addresses. Physical addresses no longer in use (e.g., 16-23), which may be referred to as “inactive physical addresses,” are illustrated with dashed formatting.
  • a column labeled “Bank” lists the memory banks 40A, 40B, and 40C (bank 0, bank 1, and bank 2), which, for explanatory purposes only, are shown to hold eight DIMM addresses.
  • the size of the memory banks 40 and their memory addresses may be much larger, and may represent any size chunk of memory that may be manipulated by the memory controller 36 (e.g., a cache line or page).
  • the memory bank 40 that is to be shut down e.g., bank 2 is illustrated with dashed formatting.
  • a column labeled "DIMM Address (3 Banks)” illustrates the original mapping of physical addresses to certain hardware memory storage locations, or DIMM addresses, on the three memory banks 40A, 40B, and 40C.
  • a column labeled "DIMM Address (2 Banks)” illustrates a remapping of the active physical addresses to certain hardware memory storage locations, or DIMM addresses, when the number of active memory banks 40 is reduced from three to two.
  • the OS running on the processor(s) 12, software running on the operating system (OS), or the processor(s) 12 may arrange the in-use physical addresses such that the in-use physical addresses are outside the area of memory to be powered down. For example, certain wired physical addresses or pages may be cleared out.
  • the OS, the software, and/or the processor(s) 12 may cause the memory controller 36 to receive a command to begin the dynamic remapping process. Additionally or alternatively, the memory controller 36 may automatically determine to begin the dynamic remapping process as discussed above.
  • the memory controller 36 may employ an alternative dynamic memory mapping function to remap the in-use physical addresses that have been mapped, in the three-bank mapping scheme, to DIMM addresses of the third memory bank 40.
  • the alternative memory mapping function may be any suitable function that maintains the mapping of the active physical addresses located on the memory banks 40A, 40B, and 40C that will remain in use, while remapping DIMM addresses that map, in the three-bank scheme, to inactive physical addresses.
  • the memory banks 0 and 1 are to remain active, and the active physical addresses (e.g., 0-15) mapped to DIMM addresses on banks 0 and 1 may not change.
  • variable BankReindex 3_2 is employed, in combination with the variables FirstBankEntries 3_2, FirstBankOffset 3_2, and SecondBankOffset 3_2 to ascertain the alternative DIMM address mapping. Used in determining these variables, the variable BankSize represents the quantity of physical addresses per memory bank 40, the variable TotalBanks represents the total number of memory banks 40A, 40B, and 40C, and the variable ReducedBanks3 _2 represents the number of memory banks 40A, 40B, and 40C remaining once the selected memory bank 40 is shut down.
  • the physical address "4" maps to DIMM address "9."
  • the physical address "4" may continue to be mapped to the DIMM address "9.”
  • the physical address "2" may be remapped from the DIMM address "16,” which is located on the memory bank 40 that is to be shut down (e.g., bank 2), to the DIMM address "6.”
  • the memory controller 36 may employ a copy function 122 to copy the data from the soon-to-be-shut-down memory bank 40 (e.g., bank 2) into the newly remapped memory banks 0 and 1. In copying the memory data, the memory controller 36 may simply write over the data in the remapped addresses of the banks 0 and 1. By way of example, the memory controller 36 may read data from the DIMM addresses of the soon-to-be-shut-down memory bank 40 that correspond to the remaining active physical addresses. The memory controller 36 may then write the data to either the alternative DIMM address or both the alternative DIMM address and the original DIMM address.
  • a copy function 122 to copy the data from the soon-to-be-shut-down memory bank 40 (e.g., bank 2) into the newly remapped memory banks 0 and 1.
  • the memory controller 36 may simply write over the data in the remapped addresses of the banks 0 and 1.
  • the memory controller 36 may read data from the DIMM addresses of the soon-to-be-shut-down memory bank 40 that
  • the copy function 122 may efficiently transfer the data from bank 2 to banks 0 and 1 without other intermediate copying steps, such as copying from any of the active banks 0 or 1.
  • the memory remapping process that may take place as shown in the memory remapping diagram 120 may not involve any action on the part of a translation look aside buffer (TLB) of the MMU 39 of the processor(s) 12, but rather may take place dynamically with only the memory controller 36 performing remapping and copying in a rapid, efficient manner.
  • TLB translation look aside buffer
  • the memory bank 40 to be shut down may continue to be accessible while the remapping is taking place. Since the data stored on the memory bank 40 to be shut down (e.g., bank 2) is written into both the three-bank DIMM address mapping and two-bank DIMM address mapping (e.g., written to hardware memory locations on banks 0, bank 1, and bank 2), the data on the memory bank 40 to be shut down will remain accessible at the three-bank DIMM address mapping until the memory bank 40 is actually shut down. Moreover, it should be understood that this principal of remapping may be extended to any reduced number of memory banks 40, and not only the case from three memory banks 40 to two memory banks 40.
  • the memory controller 36 may continue to operate in the two-bank DIMM address mapping. Because the memory controller 36 continues to operate in the two-bank DIMM address mapping, the remapping operation may be aborted at any time before the memory bank to be shut down (e.g., bank 2) is finally shut down.
  • the memory controller 36 may swap, rather than destructively copy, the data from the soon-to-be-shut-down memory bank 40 (e.g., bank 2) into the newly remapped memory banks 0 and 1.
  • physical address 2 may be swapped with the physical address 18, the physical address 5 with 21, and so forth.
  • swapping may enable the data from physical addresses not currently in use to be available in the future.
  • the memory bank 40 to be shut down may be placed into a self-refresh mode, which may consume relatively little power but may preserve the data stored within the memory bank 40.
  • the originally swapped physical addresses may be swapped back.
  • the original 3-bank memory mapping function of Equations (1) and (2) may be calculated to distribute the physical memory addresses across the DIMM addresses such that each memory bank 40 includes fewer related physical addresses than the maximum possible without latency. Thereafter, using the alternative mapping function of Equations (3)-(7), the memory bank 40 may not incur additional latency, even if two related physical addresses are mapped to the same memory bank 40. Thus, two related physical addresses may be mapped to the same memory bank 40 based on the alternative mapping function, but may not significantly increase the latency of the memory 14.
  • a flowchart 124 describes an embodiment of a method for carrying out the process shown in the memory remapping diagram 120.
  • the memory management system 34 may initiate entrance to a two-bank memory mapping mode.
  • the memory controller 36 may employ the alternative memory mapping function of Equations (3)-(7) when physical addresses are received from the processor(s) 12.
  • the memory data of the remaining active physical addresses on memory bank 2 may be copied onto the newly remapped memory addresses of memory banks 0 and 1, as illustrated in the memory mapping diagram 120.
  • the memory controller 36 may only read data from the DIMM addresses of the memory bank 40 to be shut down, but may write the data back to the three-bank DIMM addresses and to the two-bank DIMM addresses. While such copying takes place, the memory controller 36 may continue to operate in the three-bank memory mapping mode, asynchronously transitioning toward the two-bank memory mapping mode, provided each copy is atomic (i.e., each read of a given DIMM address is accompanied by writes back to the two DIMM addresses). That is, from the point of view of the operating system (OS), the memory controller 36 appears to be operating in the two-bank memory mapping. From the perspective of the OS, the OS is merely employing a reduced address space.
  • OS operating system
  • the memory controller 36 may carry out these techniques in a manner unseen by the processor(s) 12 or operating system (OS).
  • OS may cause the memory controller 36 to perform certain or all of the blocks 126, 128, and 130 (e.g., the OS may cause atomic copies in block 130).
  • the variable DIMMAddrBank represents an offset value.
  • the variable BankSize represents the quantity of physical addresses per memory bank 40
  • the variable TotalBanks represents the total number of memory banks 40A, 40B, and 40C
  • the variable TotalMemory represents the total number of DIMM addresses.
  • the DIMM address "16" on bank 2 maps to the physical address "2," in accordance with Equations (8) and (9) above. Since the DIMM address "16" holds data associated with the physical address "2," which is among the physical addresses to remain active, the memory controller 36 may read the DIMM address "16" and copy the data stored therein onto the alternative DIMM address mapping for the physical address "2" as well as the original DIMM address mapping for the physical address "2.” The memory controller 36 may continue to assess each of the DIMM addresses of the memory bank 40 to be shut down (e.g., bank 2) until an inactive physical address (e.g., "17") is determined.
  • an inactive physical address e.g., "17
  • the memory bank 40 to be shut down may continue to be accessible while the remapping is taking place. Since the data stored on the memory bank 40 to be shut down (e.g., bank 2) is destructively copied and written into both the alternative DIMM address mapping and original DIMM address mapping, the data on the memory bank 40 to be shut down will remain accessible at that original DIMM address mapping until the memory bank 40 is actually shut down, at which the memory controller 36 may follow the alternative DIMM address mapping scheme. It should further be appreciated that this principal of remapping may be extended to any reduced number of memory banks 40.
  • one or memory banks 40 may be inactive.
  • the previously shut down memory bank 40 e.g., bank 2
  • data may be read from the DIMM address defined by a two-bank memory mapping, but may be written to the DIMM addresses defined by a three-bank memory mapping and the two-bank memory mapping.
  • the third bank may be inactive at least some of the time (e.g., in self-refresh mode while not being written to), saving power.
  • the memory controller 36 switches back to the three-bank mapping, the copying of data back to the three-bank mapping of DIMM addresses will have already occurred, saving time.
  • Such an embodiment may be particularly useful for asymmetric, read-dominated circumstances.
  • FIGS. 24 and 25 illustrate the alternative memory mapping 120, which is described in greater detail above with reference to FIG. 10 . As such, this discussion is not reproduced here.
  • FIGS. 24 and 25 only bank 0 and bank 1 are active and the memory controller 36 is generally operating in a two-bank memory mapping. That is, in the example shown in FIG. 24 , when the memory controller 36 opts to perform a read operation 244 to read the physical address "2," the memory controller only reads from the DIMM address "6" that maps to the physical address "2" in the two-bank memory mapping.
  • the memory bank 2 may remain inactive when data is being read, saving power.
  • the memory controller 36 may write the data not only to the two-bank memory mapping, but also to the three-bank memory mapping. To do so, the memory bank 2 may be activated at least temporarily.
  • the memory controller 36 when the memory controller 36 opts to perform a write operation 246 to write the physical address "2," the memory controller may write not only to the DIMM address "6" on bank 0, but also to the DIMM address "16" on bank 2.
  • the memory controller 36 later transitions from the two-bank memory mapping to the three-bank memory mapping, the data from the DIMM address "6" need not be copied to the DIMM address "16," because the content of the physical address "2" already is located in the DIMM address "16.”
  • the process of dynamic memory remapping illustrated in FIG. 11 may be extended to reduce the amount of active memory further, as shown by a memory mapping diagram 140 of FIG. 13 .
  • the memory mapping diagram 140 illustrates a reduction in memory use from two memory banks 40A, 40B, and 40C to one memory bank 40.
  • a column labeled "Physical Address” lists physical addresses. Physical addresses no longer in use (e.g., 8-24), which may be referred to as "inactive physical addresses,” are illustrated with dashed formatting.
  • a column labeled "Bank” lists the memory banks 40A, 40B, and 40C, each of which, for explanatory purposes only, hold eight DIMM addresses.
  • the size of the memory banks 40 and their memory addresses may be much larger, and may represent any size chunk of memory that may be manipulated by the memory controller 36 (e.g., a cache line or page).
  • the memory bank 40 that is to be shut down e.g., bank 1
  • the memory bank 40 previously shut down are illustrated with dashed formatting.
  • a column labeled "DIMM Address (2 Banks)” illustrates the two-bank memory mapping scheme of FIG. 11 .
  • a column labeled "DIMM Address (1 Bank)” illustrates a remapping of the active physical addresses to certain DIMM addresses, when the number of active memory banks 40A, 40B, and 40C is reduced from two to one.
  • the memory controller 36 may carry out the memory mapping procedure outlined in the memory mapping diagram 140 in a similar manner to the methods discussed above with reference to FIGS. 10 and 12 .
  • the memory controller 36 may employ another alternative dynamic memory mapping function to remap the in-use physical addresses that have been mapped, in the two-bank mapping scheme, to DIMM addresses of the second memory bank 40, which will be shut down.
  • the alternative memory mapping function may be any suitable function that maintains the mapping of the active physical addresses located on the memory bank 40 that will remain in use, while remapping DIMM addresses that map, in the two-bank scheme, to inactive physical addresses.
  • the memory bank 0 is to remain active, and those of the active physical addresses (e.g., 0-7) mapped to DIMM addresses on bank 0 may not change.
  • the memory controller 36 may remap these DIMM addresses on-the-fly.
  • variable BankReindex2 _1 is employed, in combination with the variables FirstBankEntries 2_1, and FirstBankOffset 2_1 to ascertain the alternative DIMM address mapping. Used in determining these variables, the variable BankSize represents the quantity of physical addresses per memory bank 40, the variable TotalBanks represents the total number of memory banks 40A, 40B, and 40C, and the variable ReducedBanks2 _1 represents the number of memory banks 40A, 40B, and 40C remaining once the selected memory bank 40 (e.g., bank 1) is shut down.
  • the selected memory bank 40 e.g., bank 1
  • the physical address "4" may still map to DIMM address "9.” Based on Equations (10)-(13), as shown in FIG. 11 , the physical address "4" may be remapped to DIMM address "4" in the one-bank memory mapping scheme. Similarly, the physical address "2,” mapped to DIMM address "6" in the two-bank memory mapping scheme, may remain mapped to DIMM address "6" in the one-bank memory mapping scheme.
  • the memory controller 36 may perform a copy function 142 to transfer the remaining active memory addresses to the remapped addresses in the 1-bank memory mapping scheme.
  • the copy function 142 may take place in a manner similar to the copy function 122 discussed above.
  • the memory shut-down process may, in certain embodiments, be employed by the electronic device 10 for power management and/or thermal management under specific circumstances.
  • FIG. 14 illustrates a flowchart 150 describing an embodiment of such a method for power management and/or thermal management.
  • the electronic device 10 may enter an idle state or may exceed a temperature threshold.
  • the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS may determine that power consumption should be reduced to save power and/or to prevent additional heat from being generated.
  • the OS may reduce memory usage using any of a variety of measures, which may include limiting the frame rate of graphics on the display 18, consolidating the working set of memory to fit into smaller areas of memory, and/or reducing unbound memory. Additionally or alternatively, the electronic device 10 may enter an idle or reduced-memory-usage state when the power source 26 switches from AC power to battery power, which may involve automatically reducing the video frame rate. In certain embodiments, the electronic device 10 may reduce memory usage by backing up memory data onto the nonvolatile storage 16, thereafter indicating that the higher order physical addresses to which the data corresponds is no longer in use.
  • the memory controller 36 may perform one of the memory remapping procedures described above and shut down a portion of the memory 14 (e.g., bank 2), conserving power and/or reducing heat.
  • a portion of the memory 14 e.g., bank 2
  • certain DIMM addresses on the memory bank 40 that is to be shut down e.g., bank 2 may be put to use as cache before that memory bank 40 is shut down.
  • FIGS. 15 and 16 illustrate memory remapping diagrams for remapping, on-the-fly, from fewer memory banks 40A, 40B, and 40C to more memory banks 40A, 40B, and 40C.
  • FIG. 15 a memory remapping diagram 160 illustrates a process for transitioning to the two-bank memory mapping scheme from the one-bank memory mapping scheme.
  • a column labeled "Physical Address" lists physical addresses.
  • a column labeled “Bank” lists the memory banks 40A, 40B, and 40C, which, for explanatory purposes only, each hold eight DIMM addresses. In practice, it should be understood that the size of the memory banks 40 and their memory addresses may be much larger, and may represent any size chunk of memory that may be manipulated by the memory controller 36 (e.g., a cache line or page).
  • a column labeled “DIMM Address (2 Banks)” illustrates the two-bank memory mapping scheme, and a column labeled "DIMM Address (1 Bank)” illustrates the one-bank memory mapping scheme.
  • the memory controller 36 may perform a copy function 162.
  • the memory controller 36 may determine which DIMM addresses of bank 0 currently store data associated with physical addresses that, in the two-bank memory mapping scheme, map to DIMM addresses of bank 1.
  • the memory controller 36 may read from these DIMM addresses and may write their contents onto at least the DIMM addresses associated with these physical addresses of the two-bank memory mapping scheme. In certain embodiments, the memory controller 36 may write such data onto both the DIMM addresses of the one-bank memory mapping scheme and of the two-bank memory mapping scheme.
  • a memory remapping diagram 170 illustrates a process for transitioning to the three-bank memory mapping scheme from the two-bank memory mapping scheme.
  • a column labeled "Physical Address” lists physical addresses.
  • a column labeled "Bank” lists the memory banks 40A, 40B, and 40C, which, for explanatory purposes only, each hold eight DIMM addresses. In practice, it should be understood that the size of the memory banks 40 and their memory addresses may be much larger, and may represent any size chunk of memory that may be manipulated by the memory controller 36 (e.g., a cache line or page).
  • a column labeled "DIMM Address (3 Banks)" illustrates the three-bank memory mapping scheme, and a column labeled "DIMM Address (2 Banks)” illustrates the two-bank memory mapping scheme.
  • the memory controller 36 may perform a copy function 172.
  • the memory controller 36 may determine which DIMM addresses of banks 0 and 1 currently store data associated with physical addresses that, in the three-bank memory mapping scheme, map to DIMM addresses of bank 2.
  • the memory controller 36 may read from these DIMM addresses and may write their contents onto at least the DIMM addresses associated with these physical addresses of the three-bank memory mapping scheme. In certain embodiments, the memory controller 36 may write such data onto both the DIMM addresses associated with these physical addresses for both the two-bank memory mapping scheme and the three-bank memory mapping scheme.
  • the memory controller 36 may have swapped, rather than destructively copied, the data from the shut-down memory bank 40 (e.g., bank 2 or bank 1) into the newly remapped memory banks 0 and/or 1.
  • the memory bank 40 that was shut down may have been placed in a low-power self-refresh mode. As such, the shut-down memory bank 40 may have preserved the data stored in its physical addresses.
  • This data may be recovered when the shut-down memory bank 40 is reactivated and remapped (e.g., as according to FIGS. 15 or 16 ).
  • such data may instead be swapped.
  • performing such a swap function rather than a destructive copy function may involve additional memory buffer(s) to hold the physical addresses being swapped while the swap is taking place, swapping may enable the previously inaccessible data to be accessible again without need for restoring such data from non-volatile storage 16.
  • the electronic device 10 may balance performance needs with memory usage needs to reduce the amount of power employed by the electronic device 10 in certain situations, as illustrated in a flowchart 180 of FIG. 17 .
  • the memory controller 36, the processor(s) 12, the OS running on the processor(s) 12, and/or software running on the OS may assess memory usage and performance needs of the electronic device 10. This may be accomplished, for example, by monitoring the amount of memory currently in use and/or using performance counters tracked by the electronic device 10.
  • decision blocks 184 and 186 if performance needs are low and memory needs are low, the electronic device 10 may undertake the remapping procedures to power down one or more of the memory banks 40A, 40B, and 40C according to the techniques described above in block 188.
  • the electronic device 10 may reduce the clock frequencies of the memory banks 40A, 40B, and 40C and/or the memory buses 42, 44, and 46 in block 190. Additionally or alternatively, the electronic device 10 may perform the system idle procedure described above with reference to FIG. 14 to reduce the amount of memory in use, before performing one of the remapping and memory shut-down techniques described above.
  • the memory 14 may be remapped and an additional memory bank 40A, 40B, or 40C may be reactivated, as shown by block 194. Though not all of the additional DIMM addresses may be used by the electronic device 10, by distributing the memory data across additional memory banks 40, the memory access bandwidth may be increased.
  • the memory controller 36 may perform the remapping procedure in a manner such as described above with reference to FIGS. 15 and/or 16. If performance needs are high and memory usage is high, as indicated by decision blocks 184 and 192, all memory banks 40A, 40B, and 40C may be made active and in use, as shown by block 196.
  • Performance may be improved further by reducing the number of certain types of accesses to the memory banks 40.
  • the DIMM addresses may be accessible via a row address selection (RAS), a column address selection (CAS), and a device selection (DEV). Since changing memory rows through changes in RAS may produce greater latency than changing memory columns through changes in CAS, on-the-fly memory remapping techniques may be employed that reduce the number of RAS changes and increase the number of CAS changes. Techniques for performing such memory remapping are discussed in greater detail below.
  • the memory controller 36 may remap certain address bits when certain memory is requested by the processor(s) 12.
  • a cache line address 200 that corresponds to an address space of 256 cache-line addresses may include 8 bits, labeled A0 to A7 in order of significance. It should be appreciated that, in practice, cache line addresses may be longer or shorter. Indeed, the present embodiments illustrated in FIGS. 18, 19A-C , 21A-E, and 22A-B are shown to include 8 bits to encode an address space of 256 cache lines for ease of explanation only.
  • copying of data between DRAM devices should be reduced. For example, if one-half the memory of the memory banks 40 is to be powered down, then no more than one-half the memory should be relocated when the powering down occurs.
  • the electronic device 10 may operate in 3 different configurations, in which either 1, 2, or 4 memory banks 40 remains powered, and where each configuration doubles the amount of memory available (e.g., each memory bank 40 is of equal size).
  • each configuration doubles the amount of memory available (e.g., each memory bank 40 is of equal size).
  • unnecessary copying operations may be avoided by applying three different address-mapping schemes. These schemes may enable a high degree of memory bank 40 interleaving in each mode, while reducing row changes regardless of the mode.
  • the memory controller 36 may map the cache line address 200 in a manner that limits mapping changes to only two bits each time the configuration changes.
  • FIGS. 19A-C represent mappings for a configuration starting with four memory banks 40 of equal size.
  • a cache line address 200 that corresponds to an address space of 256 cache-line addresses may include 8 bits, labeled A0 to A7 in order of significance. It should be appreciated that, in practice, cache line addresses may be longer or shorter.
  • the present embodiments illustrated in FIGS. 18, 19A-C , 21A-E, and 22A-B are shown to include 8 bits to encode an address space of 256 cache lines for ease of explanation only.
  • the cache line address 200 of FIGS. 21A-C each map 64 cache addresses.
  • FIG. 19A represents a configuration when all four memory banks 40 are active
  • FIG. 19B represents a configuration when two of the memory banks 40 are active
  • FIG. 19C represents a configuration when only one of the memory banks 40 is active.
  • R0-R2 correspond to RAS bits
  • C0-C2 correspond to CAS bits
  • D0 and D1 correspond to DEV bits.
  • the system memory map for the mode may map cache addresses 128-255 to the two fastest memory banks 40 and cache addresses 0-63 and 64-127 respectively to slower memory banks 40.
  • system memory map for the mode may map cache addresses 0-63 and 64-127 respectively to slower memory banks 40, and when only one of the four memory banks 40 is active, the system memory map for the mode may map cache addresses 0-63 to the slowest of the memory banks 40.
  • one of the bits may be relocated to a portion of the system address-map that will be powered-down. Since the DEV bits D0 and D1 reside in the lower-ordered bits to achieve maximum memory bank 40 interleaving when all four memory banks 40 are active ( FIG. 19A ), and the powering-down of a memory bank 40 implies that that memory bank 40 may no longer be accessed, one of the DEV bits may be relocated from the lower-ordered bits to one of the higher-ordered bits that will no longer be accessed ( FIG. 19B ), in order to cause accesses to the disabled memory bank 40 not to occur.
  • the bit that should be exchanged with a DEV bit should be a CAS bit, implying that the CAS bit formerly used the higher-ordered bits of the cache address.
  • RAS must change more frequently during streaming with multiple memory banks 40 enabled that might otherwise be achieved, but under no circumstances should it change more frequently than when only a single memory bank 40 is enabled.
  • cache line address bits A1 and A7 are swapped.
  • cache line address bits A1 and A6 are swapped.
  • FIGS. 19A-C result in the following translation between cache line addresses and system addresses, in which the notation ([#],[#],[#]) [#] corresponds to ([RAS], [CAS], [DEV]) [system address]: Table 1.
  • Cache Address four-memory-bank configuration FIG. 19A
  • two-memory-bank configuration FIG. 19B
  • one-memory-bank configuration FIG.
  • the electronic device 10 may have only 3 memory banks 40, one memory bank 40 of which may have more memory than the remaining two memory banks 40.
  • a faster memory bank 40 may have the same capacity as the combined capacity of the two slower memory banks 40 (e.g., the fast memory bank 40 may have twice the capacity of each of the two remaining slow memory devices 40).
  • the above design strategy can still be employed by mapping DEV[0] (D0) to the additional CAS bit (e.g., CAS[3] (C3)) on the faster of the memory banks 40.
  • the electronic device 10 may follow a process such as disclosed by a flowchart 210 of FIG. 20 .
  • the flowchart 210 may begin when the electronic device 10 is operating in a full memory usage mode (e.g., if four memory banks 40 are present, all four memory banks 40 may be active). Accordingly, when the flowchart 210 begins, the memory controller 36 may map the memory banks 40 according to a scheme such as illustrated in FIG. 19A .
  • the electronic device 10 may prepare to power down half of the memory 14 available to the electronic device 10 (block 212) using any suitable technique, as discussed in greater detail above. Since faster memory banks 40 typically may consume greater amounts of power, the electronic device 10 generally may initially choose to power down the fastest memory bank(s) 40.
  • the memory controller 36 may remap the cache line addresses 200 (block 214) according to a scheme that only changes two bits of the cache line address 200 (e.g., as illustrated in FIG. 19B ). Thereafter, the memory controller 36 may copy certain data from the memory bank(s) 40 that will be powered down to the memory banks 40 that will remain active, in the manners discussed above, before powering down the chosen memory bank(s) 40 (block 216).
  • the electronic device 10 may determine to power down half of the remaining memory 14 (block 218) using any suitable technique, such as discussed above. Since faster memory banks 40 typically may consume greater amounts of power, the electronic device 10 generally may choose to power down the faster of the remaining memory bank(s) 40. By way of example, if the electronic device 10 has a total of four memory banks 40 of equal size, only the slowest memory bank 40 may be selected to remain active.
  • the memory controller 36 may next remap the cache line addresses 200 (block 220) according to a scheme that only changes two bits of the cache line address 200 (e.g., as illustrated in FIG. 19C ). Thereafter, the memory controller 36 may copy certain data from the memory bank(s) 40 that will be powered down to the memory bank(s) 40 that will remain active, in the manners discussed above, before powering down the chosen memory bank(s) 40 (block 222).
  • the electronic device 10 may employ a technique in which each mode adds or subtracts equal amounts of memory from system memory.
  • each mode adds or subtracts equal amounts of memory from system memory.
  • certain embodiments of the electronic device 10 may have 3 memory banks 40 of equal capacity. To reduce copying, only two address-mapping bits may be changed between modes in a manner similar to those discussed above.
  • FIGS. 21A-E represent mappings for a configuration starting with three memory banks 40 of equal size.
  • FIGS. 21A-C represent certain variations of a configuration when all three memory banks 40 are active
  • FIG. 21D represents a configuration when two of the memory banks 40 are active
  • FIG. 21E represents a configuration when only one of the memory banks 40 is active.
  • R0-R2 correspond to RAS bits
  • C0-C2 correspond to CAS bits
  • D0 and D1 correspond to DEV bits.
  • a cache line address 200 that corresponds to an address space of 256 cache-line addresses may include 8 bits, labeled A0 to A7 in order of significance.
  • cache line addresses may be longer or shorter. Indeed, the present embodiments illustrated in FIGS. 18, 19A-C , 21A-E, and 22A-B are shown to include 8 bits to encode an address space of 256 cache lines for ease of explanation only. Moreover, for ease of explanation only, the cache line address 200 of FIGS. 21A-C each map 64 cache addresses. That is, in certain embodiments, when all three memory banks 40 are active, the system memory map for the mode may map cache addresses 128-191 to the fastest memory bank 40 and cache addresses 0-63 and 64-127 respectively to slower memory banks 40.
  • the system memory map for the mode may map cache addresses 0-63 and 64-127 respectively to slower memory banks 40, and when only one of the three memory banks 40 is active, the system memory map for the mode may map cache addresses 0-63 to the slowest of the memory banks 40.
  • FIG. 19A Unlike the configuration of FIGS. 19A-C , in which a single mapping ( FIG. 19A ) applies for all of the cache addresses in the four-memory-bank configuration, when all three memory banks 40 are active, a different mapping applies depending on the states of certain cache line address 200 bits.
  • FIGS. 21A-E result in the following translation between cache line addresses and system addresses, in which the notation ([#],[#],[#]) [#] corresponds to ([RAS], [CAS], [DEV]) [system address]: Table 2.
  • Cache Address three-memory-bank configuration FIGS. 21A-E ) two-memory-bank configuration ( FIG. 21B ) one-memory-bank configuration ( FIG.
  • the electronic device 10 may follow a process such as disclosed by a flowchart 230 of FIG. 23 .
  • the flowchart 230 may begin when the electronic device 10 is operating in a full memory usage mode (e.g., if three memory banks 40 of equal size are present, all three memory banks 40 may be active). Accordingly, when the flowchart 230 begins, the memory controller 36 may map the memory banks 40 according to a scheme such as illustrated in FIGS. 21A-C , depending on the state of cache line address 200 bits A7 and A1.
  • the electronic device 10 may prepare to power down one-third of the memory 14 available to the electronic device 10 (block 232) using any suitable technique, as discussed in greater detail above. Since faster memory banks 40 typically may consume greater amounts of power, the electronic device 10 generally may initially choose to power down the fastest memory bank(s) 40.
  • the memory controller 36 may remap the cache line addresses 200 (block 234) according to a scheme that only changes two bits of the cache line address 200 (e.g., as illustrated in FIG. 21D and as demonstrated above). Thereafter, the memory controller 36 may copy certain data from the memory bank(s) 40 that will be powered down to the memory banks 40 that will remain active, in the manners discussed above, before powering down the chosen memory bank(s) 40 (block 236).
  • the electronic device 10 may determine to power down half of the remaining memory 14 such that only one-third of the total memory 14 remains (block 218) using any suitable technique, such as those discussed above. Since faster memory banks 40 typically may consume greater amounts of power, the electronic device 10 generally may choose to power down the faster of the remaining memory bank(s) 40. By way of example, if the electronic device 10 has a total of three memory banks 40 of equal size, only the slowest memory bank 40 may be selected to remain active.
  • the memory controller 36 may next remap the cache line addresses 200 (block 220) according to a scheme that only changes two bits of the cache line address 200 (e.g., as illustrated in FIG. 21E ). Thereafter, the memory controller 36 may copy certain data from the memory bank(s) 40 that will be powered down to the memory bank(s) 40 that will remain active, in the manners discussed above, before powering down the chosen memory bank(s) 40 (block 222).
  • copying may be completed by the memory controller 36 in a manner that is hidden from the operating system (OS) or software of the electronic device 10, or copying may be completed in manner that is controlled by the OS or software, provided each copy is atomic (e.g., a read from a first DIMM address at a first memory mapping may be immediately followed by writes to DIMM addresses of both the first and second memory mapping).
  • OS operating system
  • each copy is atomic (e.g., a read from a first DIMM address at a first memory mapping may be immediately followed by writes to DIMM addresses of both the first and second memory mapping).
EP11717074.6A 2010-04-13 2011-04-11 Memory controller mapping on-the-fly Active EP2558939B1 (en)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
US32375310P 2010-04-13 2010-04-13
US12/895,689 US8799553B2 (en) 2010-04-13 2010-09-30 Memory controller mapping on-the-fly
PCT/US2011/031885 WO2011130141A1 (en) 2010-04-13 2011-04-11 Memory controller mapping on-the-fly

Publications (2)

Publication Number Publication Date
EP2558939A1 EP2558939A1 (en) 2013-02-20
EP2558939B1 true EP2558939B1 (en) 2020-01-01

Family

ID=44761746

Family Applications (1)

Application Number Title Priority Date Filing Date
EP11717074.6A Active EP2558939B1 (en) 2010-04-13 2011-04-11 Memory controller mapping on-the-fly

Country Status (6)

Country Link
US (3) US8799553B2 (ko)
EP (1) EP2558939B1 (ko)
KR (1) KR101459866B1 (ko)
CN (1) CN102893266B (ko)
AU (1) AU2011240803B2 (ko)
WO (1) WO2011130141A1 (ko)

Families Citing this family (57)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8339626B2 (en) * 2009-09-08 2012-12-25 Samsung Electronics Co., Ltd. Image forming apparatus and controlling method thereof
US8412972B2 (en) * 2010-06-28 2013-04-02 Intel Corporation Method and apparatus for reducing power consumption for memories
US8725978B2 (en) * 2011-06-30 2014-05-13 Red Hat, Inc. Using symbol information for categorization of dynamic memory allocations
US8719539B2 (en) * 2011-06-30 2014-05-06 Red Hat, Inc. Using heuristics for field types of a structure to categorize dynamic memory allocations
US8694753B2 (en) * 2011-06-30 2014-04-08 Red Hat, Inc. Debugging a dynamic memory allocation using heuristics reliability scores
JP5877348B2 (ja) * 2011-09-28 2016-03-08 パナソニックＩｐマネジメント株式会社 メモリ制御システム及び電力制御方法
US20180144436A1 (en) * 2014-06-24 2018-05-24 Intel Corporation Reducing power for 3d workloads
US20120095607A1 (en) * 2011-12-22 2012-04-19 Wells Ryan D Method, Apparatus, and System for Energy Efficiency and Energy Conservation Through Dynamic Management of Memory and Input/Output Subsystems
JP2013164785A (ja) * 2012-02-13 2013-08-22 Nikon Corp 電子機器および電子機器の制御プログラム
JP2013222321A (ja) * 2012-04-17 2013-10-28 Sony Corp メモリ制御装置、メモリ制御方法、情報処理装置、およびプログラム
JP5624583B2 (ja) * 2012-05-31 2014-11-12 株式会社東芝 プログラム、計算処理装置、メモリ管理方法および計算機
US9256531B2 (en) 2012-06-19 2016-02-09 Samsung Electronics Co., Ltd. Memory system and SoC including linear addresss remapping logic
KR102004849B1 (ko) * 2012-08-02 2019-07-29 삼성전자 주식회사 동적 메모리 재할당 관리 방법과 상기 방법을 수행할 수 있는 장치
US9146846B2 (en) * 2012-09-14 2015-09-29 Advanced Micro Devices, Inc. Programmable physical address mapping for memory
US20140122807A1 (en) * 2012-10-31 2014-05-01 Hewlett-Packard Development Company, Lp. Memory address translations
US20140136793A1 (en) * 2012-11-13 2014-05-15 Nvidia Corporation System and method for reduced cache mode
US9218285B2 (en) * 2012-11-26 2015-12-22 Arm Limited Variable mapping of memory accesses to regions within a memory
US9720858B2 (en) 2012-12-19 2017-08-01 Nvidia Corporation Technique for performing memory access operations via texture hardware
US9348762B2 (en) 2012-12-19 2016-05-24 Nvidia Corporation Technique for accessing content-addressable memory
US9697006B2 (en) 2012-12-19 2017-07-04 Nvidia Corporation Technique for performing memory access operations via texture hardware
US9760149B2 (en) * 2013-01-08 2017-09-12 Qualcomm Incorporated Enhanced dynamic memory management with intelligent current/power consumption minimization
US9342443B2 (en) 2013-03-15 2016-05-17 Micron Technology, Inc. Systems and methods for memory system management based on thermal information of a memory system
KR102044478B1 (ko) * 2013-04-22 2019-11-13 삼성전자주식회사 드라이버 및 이를 포함하는 메모리 컨트롤러
US9043569B2 (en) 2013-05-31 2015-05-26 International Business Machines Corporation Memory data management
US9472248B2 (en) * 2014-03-28 2016-10-18 Intel Corporation Method and apparatus for implementing a heterogeneous memory subsystem
US9722335B2 (en) * 2014-05-05 2017-08-01 Qualcomm Incorporated Dual in line memory module (DIMM) connector
US9367442B2 (en) * 2014-07-12 2016-06-14 Lenovo Enterprise Solutions (Singapore) Pte. Ltd. Allocating memory usage based on voltage regulator efficiency
US20160070503A1 (en) * 2014-09-10 2016-03-10 Kabushiki Kaisha Toshiba Memory system
KR102347657B1 (ko) * 2014-12-02 2022-01-06 삼성전자 주식회사 전자 장치 및 이의 공유 캐시 메모리 제어 방법
US9678872B2 (en) 2015-01-16 2017-06-13 Oracle International Corporation Memory paging for processors using physical addresses
US10019358B2 (en) * 2015-03-20 2018-07-10 Vixs Systems Inc. Bank address remapping to load balance memory traffic among banks of memory
US20160349817A1 (en) * 2015-05-29 2016-12-01 Intel Corporation Power protected memory with centralized storage
KR102420897B1 (ko) * 2016-03-17 2022-07-18 에스케이하이닉스 주식회사 메모리 모듈, 이를 포함하는 메모리 시스템 및 그의 동작 방법
US10310975B2 (en) * 2016-05-11 2019-06-04 Seagate Technology Llc Cache offload based on predictive power parameter
US10198216B2 (en) * 2016-05-28 2019-02-05 Advanced Micro Devices, Inc. Low power memory throttling
EP3252651A1 (en) 2016-05-30 2017-12-06 Samsung Electronics Co., Ltd Computing system having an on-the-fly encryptor and an operating method thereof
US10521876B2 (en) * 2017-04-17 2019-12-31 Intel Corporation Deferred geometry rasterization technology
CN109086219B (zh) * 2017-06-14 2022-08-05 北京忆恒创源科技股份有限公司 去分配命令处理方法及其存储设备
US20190073020A1 (en) * 2017-09-01 2019-03-07 Intel Corporation Dynamic memory offlining and voltage scaling
CN107885669B (zh) * 2017-11-09 2021-06-04 上海华力微电子有限公司 一种分布式存储区块访问电路
US11119915B2 (en) 2018-02-08 2021-09-14 Samsung Electronics Co., Ltd. Dynamic memory mapping for neural networks
KR102540964B1 (ko) 2018-02-12 2023-06-07 삼성전자주식회사 입출력 장치의 활용도 및 성능을 조절하는 메모리 컨트롤러, 애플리케이션 프로세서 및 메모리 컨트롤러의 동작
KR102586741B1 (ko) * 2018-04-23 2023-10-11 에스케이하이닉스 주식회사 메모리 시스템 및 그것의 동작방법
US11256696B2 (en) 2018-10-15 2022-02-22 Ocient Holdings LLC Data set compression within a database system
US11307996B2 (en) * 2018-11-30 2022-04-19 Intel Corporation Hardware unit for reverse translation in a processor
US10936507B2 (en) * 2019-03-28 2021-03-02 Intel Corporation System, apparatus and method for application specific address mapping
US11487339B2 (en) * 2019-08-29 2022-11-01 Micron Technology, Inc. Operating mode register
US11755475B2 (en) * 2020-01-31 2023-09-12 Dell Products L.P. System and method for utilizing enhanced thermal telemetry for differential storage of data on a memory module
CN113360423A (zh) * 2020-03-03 2021-09-07 瑞昱半导体股份有限公司 数据储存系统及操作数据储存系统的方法
US11675413B2 (en) * 2020-09-30 2023-06-13 Dell Products L.P. Reducing power consumption of memory devices at an information handling system
US11604744B2 (en) 2020-10-16 2023-03-14 Alibaba Group Holding Limited Dual-modal memory interface controller
WO2022109898A1 (en) * 2020-11-26 2022-06-02 Micron Technology, Inc. Programming video data to different portions of memory
US11868266B2 (en) * 2021-03-11 2024-01-09 Micron Technology, Inc. Bank redistribution based on power consumption
WO2022193120A1 (en) * 2021-03-16 2022-09-22 Micron Technology, Inc. Logical-to-physical mapping compression techniques
US11693585B2 (en) * 2021-04-26 2023-07-04 Apple Inc. Address hashing in a multiple memory controller system
US11972140B2 (en) 2021-04-26 2024-04-30 Apple Inc. Hashing with soft memory folding
US11803471B2 (en) 2021-08-23 2023-10-31 Apple Inc. Scalable system on a chip

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030023825A1 (en) * 2001-07-30 2003-01-30 Woo Steven C Consolidation of allocated memory to reduce power consumption
US20050125701A1 (en) * 2003-12-03 2005-06-09 International Business Machines Corporation Method and system for energy management via energy-aware process scheduling
US7100013B1 (en) * 2002-08-30 2006-08-29 Nvidia Corporation Method and apparatus for partial memory power shutoff
US20080005516A1 (en) * 2006-06-30 2008-01-03 Meinschein Robert J Memory power management through high-speed intra-memory data transfer and dynamic memory address remapping
US20080059820A1 (en) * 2006-08-29 2008-03-06 Vaden Thomas L Method of reducing power consumption of a computing system by evacuating selective platform memory components thereof
US20080301474A1 (en) * 2005-12-23 2008-12-04 Nxp B.V. Performance Analysis Based System Level Power Management
US20090300299A1 (en) * 2008-05-30 2009-12-03 Saxe Eric C Dynamic Interleaving

Family Cites Families (36)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH04230508A (ja) * 1990-10-29 1992-08-19 Internatl Business Mach Corp <Ibm> 低電力消費メモリ装置
US5428758A (en) 1991-05-10 1995-06-27 Unisys Corporation Method and system for remapping memory from one physical configuration to another physical configuration
US5479624A (en) * 1992-10-14 1995-12-26 Lee Research, Inc. High-performance interleaved memory system comprising a prime number of memory modules
AU3313795A (en) * 1994-10-14 1996-04-26 Compaq Computer Corporation Circuit for placing a cache memory into low power mode in response to special bus cycles
US5956744A (en) 1995-09-08 1999-09-21 Texas Instruments Incorporated Memory configuration cache with multilevel hierarchy least recently used cache entry replacement
US6081463A (en) 1998-02-25 2000-06-27 Micron Technology, Inc. Semiconductor memory remapping
US6209074B1 (en) * 1998-04-28 2001-03-27 International Business Machines Corporation Address re-mapping for memory module using presence detect data
US6185654B1 (en) * 1998-07-17 2001-02-06 Compaq Computer Corporation Phantom resource memory address mapping system
US6546453B1 (en) 2000-08-31 2003-04-08 Compaq Information Technologies Group, L.P. Proprammable DRAM address mapping mechanism
US6535959B1 (en) * 2000-09-05 2003-03-18 Conexant Systems, Inc. Circuit and method for reducing power consumption in an instruction cache
US6751143B2 (en) * 2002-04-11 2004-06-15 Micron Technology, Inc. Method and system for low power refresh of dynamic random access memories
US6920523B2 (en) 2002-10-07 2005-07-19 Infineon Technologies Ag Bank address mapping according to bank retention time in dynamic random access memories
US6912616B2 (en) 2002-11-12 2005-06-28 Hewlett-Packard Development Company, L.P. Mapping addresses to memory banks based on at least one mathematical relationship
US7010656B2 (en) * 2003-01-28 2006-03-07 Intel Corporation Method and apparatus for memory management
US7752470B2 (en) * 2003-12-03 2010-07-06 International Business Machines Corporation Method and system for power management including device controller-based device use evaluation and power-state control
US7081897B2 (en) * 2003-12-24 2006-07-25 Intel Corporation Unified memory organization for power savings
US7685354B1 (en) 2004-06-30 2010-03-23 Sun Microsystems, Inc. Multiple-core processor with flexible mapping of processor cores to cache banks
US7318114B1 (en) * 2004-10-29 2008-01-08 Sun Microsystems, Inc. System and method for dynamic memory interleaving and de-interleaving
US8190809B2 (en) * 2004-11-23 2012-05-29 Efficient Memory Technology Shunted interleave for accessing plural memory banks, particularly those having partially accessed cells containing data for cache lines
US8010764B2 (en) * 2005-07-07 2011-08-30 International Business Machines Corporation Method and system for decreasing power consumption in memory arrays having usage-driven power management
KR100816690B1 (ko) * 2006-04-13 2008-03-27 주식회사 하이닉스반도체 온도 감지장치를 구비하는 반도체메모리소자
JP4209906B2 (ja) * 2006-08-02 2009-01-14 株式会社日立製作所 低消費電力メモリ管理方法及びその方法を用いた計算機
EP2132635B1 (en) 2007-03-30 2017-08-16 Rambus Inc. System including hierarchical memory modules having different types of integrated circuit memory devices
US8046597B2 (en) * 2007-08-14 2011-10-25 Dell Products L.P. System and method for managing storage device capacity use
US20090083561A1 (en) 2007-09-26 2009-03-26 Nikos Kaburlasos Dynamic power management of dimms
US8200999B2 (en) 2008-08-11 2012-06-12 International Business Machines Corporation Selective power reduction of memory hardware
US8639874B2 (en) 2008-12-22 2014-01-28 International Business Machines Corporation Power management of a spare DRAM on a buffered DIMM by issuing a power on/off command to the DRAM device
JP4832553B2 (ja) * 2009-06-16 2011-12-07 株式会社日立製作所 コピー開始タイミングを制御するシステム及び方法
US8392736B2 (en) 2009-07-31 2013-03-05 Hewlett-Packard Development Company, L.P. Managing memory power usage
US20110047316A1 (en) * 2009-08-19 2011-02-24 Dell Products L.P. Solid state memory device power optimization
US8285936B2 (en) * 2009-10-20 2012-10-09 The Regents Of The University Of Michigan Cache memory with power saving state
US8495299B2 (en) 2009-11-16 2013-07-23 Microsoft Corporation Non-blocking data transfer via memory cache manipulation
KR101639574B1 (ko) * 2009-12-30 2016-07-14 삼성전자주식회사 적응적 뱅크 어드레스를 제공하는 디스플레이 시스템 및 그것의 어드레스 맵핑 방법
US8412972B2 (en) * 2010-06-28 2013-04-02 Intel Corporation Method and apparatus for reducing power consumption for memories
US8356155B2 (en) * 2010-09-13 2013-01-15 Advanced Micro Devices, Inc. Dynamic RAM Phy interface with configurable power states
US9235500B2 (en) * 2010-12-07 2016-01-12 Microsoft Technology Licensing, Llc Dynamic memory allocation and relocation to create low power regions

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030023825A1 (en) * 2001-07-30 2003-01-30 Woo Steven C Consolidation of allocated memory to reduce power consumption
US7100013B1 (en) * 2002-08-30 2006-08-29 Nvidia Corporation Method and apparatus for partial memory power shutoff
US20050125701A1 (en) * 2003-12-03 2005-06-09 International Business Machines Corporation Method and system for energy management via energy-aware process scheduling
US20080301474A1 (en) * 2005-12-23 2008-12-04 Nxp B.V. Performance Analysis Based System Level Power Management
US20080005516A1 (en) * 2006-06-30 2008-01-03 Meinschein Robert J Memory power management through high-speed intra-memory data transfer and dynamic memory address remapping
US20080059820A1 (en) * 2006-08-29 2008-03-06 Vaden Thomas L Method of reducing power consumption of a computing system by evacuating selective platform memory components thereof
US20090300299A1 (en) * 2008-05-30 2009-12-03 Saxe Eric C Dynamic Interleaving

Also Published As

Publication number Publication date
US20140331020A1 (en) 2014-11-06
US9201608B2 (en) 2015-12-01
US20110252180A1 (en) 2011-10-13
CN102893266A (zh) 2013-01-23
KR20130010485A (ko) 2013-01-28
KR101459866B1 (ko) 2014-11-07
US9009383B2 (en) 2015-04-14
WO2011130141A1 (en) 2011-10-20
US8799553B2 (en) 2014-08-05
EP2558939A1 (en) 2013-02-20
CN102893266B (zh) 2015-06-03
AU2011240803B2 (en) 2014-05-29
AU2011240803A1 (en) 2012-11-22
US20140325173A1 (en) 2014-10-30

Similar Documents

Publication Publication Date Title
EP2558939B1 (en) Memory controller mapping on-the-fly
US11200176B2 (en) Dynamic partial power down of memory-side cache in a 2-level memory hierarchy
US7549034B2 (en) Redistribution of memory to reduce computer system power consumption
US10521003B2 (en) Method and apparatus to shutdown a memory channel
US8010764B2 (en) Method and system for decreasing power consumption in memory arrays having usage-driven power management
EP2936272B1 (en) Reducing power consumption of volatile memory via use of non-volatile memory
US10353454B2 (en) Information processing apparatus and computer program product for changing swap spaces based on a performance threshold
US20070005998A1 (en) Various apparatuses and methods for reduced power states in system memory
US7821864B2 (en) Power management of memory via wake/sleep cycles
US9471131B2 (en) Apparatus and machine for reducing power consumption of memory including a plurality of segment areas, method therefore and non-transitory computer readable medium
US20110047316A1 (en) Solid state memory device power optimization
US20030028711A1 (en) Monitoring in-use memory areas for power conservation
WO2005069148A2 (en) Memory management method and related system
US20060047493A1 (en) Memory management to enable memory deep power down mode in general computing systems
US20120017052A1 (en) Information Handling System Universal Memory Wear Leveling System and Method
US20160334861A1 (en) Power management for a data storage system
US8068373B1 (en) Power management of memory via wake/sleep cycles
US20220083280A1 (en) Method and apparatus to reduce latency for random read workloads in a solid state drive
JP2011018237A (ja) ディスクアレイ装置及びその制御方法
Brandt et al. Analysis of the PASR Standard and its usability in Handheld Operating Systems such as Linux

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20121108

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

DAX Request for extension of the european patent (deleted)
17Q First examination report despatched

Effective date: 20151208

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: EXAMINATION IS IN PROGRESS

RAP1 Party data changed (applicant data changed or rights of an application transferred)

Owner name: APPLE INC.

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: GRANT OF PATENT IS INTENDED

RIC1 Information provided on ipc code assigned before grant

Ipc: G06F 12/02 20060101AFI20190624BHEP

Ipc: G06F 1/32 20190101ALI20190624BHEP

INTG Intention to grant announced

Effective date: 20190717

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE PATENT HAS BEEN GRANTED

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

Ref country code: AT

Ref legal event code: REF

Ref document number: 1220668

Country of ref document: AT

Kind code of ref document: T

Effective date: 20200115

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: DE

Ref legal event code: R096

Ref document number: 602011064346

Country of ref document: DE

REG Reference to a national code

Ref country code: NL

Ref legal event code: MP

Effective date: 20200101

REG Reference to a national code

Ref country code: LT

Ref legal event code: MG4D

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200527

Ref country code: NO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200401

Ref country code: LT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: NL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: RS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200402

Ref country code: IS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200501

Ref country code: HR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: SE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: LV

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200401

REG Reference to a national code

Ref country code: DE

Ref legal event code: R097

Ref document number: 602011064346

Country of ref document: DE

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: ES

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: DK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: SM

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

REG Reference to a national code

Ref country code: AT

Ref legal event code: MK05

Ref document number: 1220668

Country of ref document: AT

Kind code of ref document: T

Effective date: 20200101

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MC

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

REG Reference to a national code

Ref country code: CH

Ref legal event code: PL

26N No opposition filed

Effective date: 20201002

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CH

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200430

Ref country code: IT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: LI

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200430

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200430

Ref country code: LU

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200411

Ref country code: AT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

REG Reference to a national code

Ref country code: BE

Ref legal event code: MM

Effective date: 20200430

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: PL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: BE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200430

GBPC Gb: european patent ceased through non-payment of renewal fee

Effective date: 20200411

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: GB

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200411

Ref country code: IE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20200411

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: TR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: MT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: CY

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

Ref country code: AL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20200101

P01 Opt-out of the competence of the unified patent court (upc) registered

Effective date: 20230523

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20230307

Year of fee payment: 13